emadbinabid / fly-time Goto Github PK

Currently, whenever we need to create a database, collections, or indexes, we do it manually through the shell. This process is repetitive and not very developer-friendly.

To overcome this, we need to come up with a one-off mechanism (a shell script) that would set up the database infrastructure, create relevant collections, and import data in those collections, if needed.

As of now, we have an isolated and independent Docker container for Fly Time. The ArangoDB instance it uses needs to run in the local environment and then the environment variables need to be set so that the containerized fly-time can talk to the locally-hosted ArangoDB instance.

As part of this ticket, we need to develop Helm chart that will install fly-time and also the ArangoDB instance in the distributed Kubernetes environment and assign the DB connection values from within the Kubernetes cluster.

On a higher level, we will need to create a deployment pod, corresponding Kubernetes service, a job that will create the database, and a config map.

A lot of factors can affect the whole customer journey from one airport to another. One example would be airport delays.

We need to come up with a solution that takes into account the airport delays as well when finding the connections between airports.

https://app.swaggerhub.com/apis/FAA/ASWS/1.1.0#/FAA%20Airport%20Status%20Service/getDelays looks like a useful API to look at.

For a given day, on a higher level, after querying the flights from DB, we need to query the above API to get the delays between airports and add the delay-time to returned results.

But this approach does not guarantee the cumulative lowest travel time if we label the airport delays as "travel time" as well.

The business prospect of this use-case is open for discussion.